Validation of NORM (Needs Oriented Framework for Producing Requirements Decision Material) Framework in Industry

Abstract: Context: Pre-project Requirements Engineering (RE) decisions are considered as very complex in Market Driven Requirements Engineering (MDRE). The complexity and economical impact of RE decisions motivates the need to support and understand RE decisions in MDRE. The purpose of Pre-project activities is to gather, analyze, select, and plan requirements for future releases of the software product in MDRE environment. Correct and timely pre-project decisions (for example initial requirements selection) are dependent on the quality and the availability of the provided decision material (for example requirements, business cases etc) for RE decisions. A Requirements Engineering framework called NORM (Needs Oriented Framework for Producing Requirements Decision Material) in collaboration with Ericsson AB is previously created with the intention of facilitating and speeding up the RE processes by ensuring that the right and sufficient decision material is available at the right time for pre-project RE decisions. NORM needs to be validated in software companies practicing MDRE in order to make it generally applicable in industry. Objectives: NORM framework was partially validated with informal interviews at one business unit within Ericsson AB but needed to be tested or validated in different environments (Telecom, e-commerce etc) in software industry to make sure that it becomes generally applicable in industry. The scope of this master's thesis is to investigate how to tailor NORM to improve a company's requirements engineering processes with the help of value stream mapping (VSM), and to evaluate the effort required and impact of this tailored process. Two MDRE based companies are chosen in this thesis work in order to validate NORM framework in software industry. Methods: There are two main research methods in this research work; Interviews and Document Analysis. In two MDRE based companies, we conducted interviews of employees involved in pre-project RE decision activities in a product, and accessed the requirements documents for the same product in order to initiate the NORM framework. Finally, an email was sent to the interviewees in both companies involved in RE with less than 10 questions in order to confirm our findings for validation of NORM framework. Results: With the information of initiation of NORM framework and using VSM we are able to chart the current decision points (DP) or decisions in the company’s RE process and the decision material that is used for each DP. We then have investigated what decision material is just enough, and when do a company need to produce it so that it is ready when the decision should be taken. We have looked at things such as waiting times, utilization and usefulness of produced decision material, and finally it is suggested to companies how their RE decision process in MDRE environment can be improved. Conclusions: Interviewees provided us the positive feedback on NORM, and they liked our novel research work in the form of NORM framework. Interviewees felt that the provided NORM structure of presenting pre-project decisions and its steps of identifying Appropriate Decision Material (ADM) is useful and usable, scalable and is according to industry relevance. It is concluded that the validation of NORM framework achieved positive feedback and results.